Because of the absence of any accepted method for predicting caries, it is proposed to examine the procedures previously used for that purpose, and seek new approaches to the problem. For the latter purpose the nature of the variables recently described in studies of bacteria, saliva, food, tooth structure and caries initiation will be sifted for use as caries predictors. Task Forces representing the various biologic and epidemiologic aspects of caries will select, in their field of responsibility, the most promising approaches to practical caries prediction and indicate how these can be brought to a stage of clinical reliability. The conference papers, summaries of the discussion and the Task Force reports will be edited and submitted to Information Retrieval Inc. for publication. Copies will be sent to all members of the American Association for Dental Research.